Seminar: Data-Driven Model-Free Control of Mineral Grinding Using Adaptive Dynamic Programming and Reference Governor, 8th April, 1p

When: Thursday 8th of April, 1pm AEST

Where: This seminar will be presented online, RSVP here.

Speaker: Dr Xinglong Lu

Title: Data-Driven Model-Free Control of Mineral Grinding Using Adaptive Dynamic Programming and Reference Governor

Abstract: Mineral grinding is one of the most energy consuming and costly procedures in mineral processing industry. Operation performance of mineral grinding is measured in terms of grinding product particle size and circulating load, two metrics that represent product quality and operation efficiency, respectively. However, the dynamic model of mineral grinding is difficult to establish and too complicated to use in a control design. In this talk, we present a data-driven model-free control method for mineral grinding control with input constraints. The controller is composed of two components: a steady-state internal model and an optimal regulator. A reference governor is used as the internal model to provide reference for input while dealing with input constraints and infeasible setpoint issue, where in some cases, a setpoint is not reachable under certain constraints. A lookup table embedded in the reference governor mapping steady-state outputs to inputs provides feasible reference for both output and input, which allows for a well-defined stabilizing problem. A novel policy iteration algorithm is proposed for the optimal regulator design without system modeling. Simulation results comparing performances of mineral grinding control with and without the reference governor show the effectiveness of the proposed control method.

Bio: Xinglong Lu received his PhD degree in control theory and control engineering from Northeastern University, China in 2019. From 2015 to 2016, he was a visiting student at the Department of Naval Architecture and Marine Engineering, University of Michigan, Ann Arbor, USA. Since 2019, he has been a research associate in Rio Tinto Centre for Mine Automation, the Australian Centre for Field Robotics, the University of Sydney. His research interests include nonlinear control systems, statistical process monitoring, industrial process control, and short-term planning for open-pit mines.

Contacts

Sydney Institute for Robotics and Intelligent Systems
info@acfr.usyd.edu.au